In Raju, Dhar, and Morrison (1994), a paper that appeared earlier in t
his journal, we developed an analytical model and conducted empirical
analyses to examine the effect of package coupons on market share. In
this epilogue, we extend the analytical framework in our earlier paper
to study the relative impact of package coupons on profits. We also r
eport findings from five quasi-experiments (including two new quasi-ex
periments) that provide empirical validity to our model-based predicti
ons. Our results have significance for brand managers in packaged good
s firms. Our analysis suggest that while selecting among package coupo
ns, brand managers should carefully define which performance criterion
to use: market share impact, redemption rate, or profit impact. Packa
ge coupons that lead to the highest market share impact (or redemption
s) may not lead to the highest profit impact. We compare the relative
profit impact of the following package coupons: peel-off, on-pack, and
in-pack. Peel-off coupons must be redeemed on the same purchase occas
ion on which they are obtained. On-pack coupons are obtained at one pu
rchase occasion but can only be redeemed for a discount on the coupone
d brand at a future purchase occasion. In-pack coupons are similar to
on-pack coupons except that the consumer is unaware of the presence of
these coupons when the product is purchased (in-pack coupons are prin
ted or placed inside the package). Consumers with an in-pack or on-pac
k coupon from a previous purchase occasion will have a higher probabil
ity of purchasing the couponed brand even if the brand was not current
ly offering coupons. Consequently, to understand choice behavior, it i
s not enough to take into consideration the current choice environment
; our model therefore must also keep track of whether or not a consume
r has a package coupon that was obtained on an earlier purchase occasi
on. In other words, since choice is dependent on the purchase environm
ent as well as the state of the consumer, we use a Markov model to rep
resent the choice process. Analytical results are derived based on the
long run probabilities of the Markov transition matrix. Our analytica
l and empirical results suggest that by and large, of the various pack
age coupons examined in our research, on-pack coupons lead to the high
est impact on profits. Furthermore, while peel-offs lead to a higher m
arket share than in-packs, because in-packs stimulate repurchase among
previous buyers, they lead to higher profits than peel-offs; though o
nly for stronger brands.